The tables consisted of a piece of plexiglass atop a wooden frame, covered by a white tarp canopy which was open at the two narrow ends of the table. The canopy enclosed the table so there were only two open ends as well as prevented glare on the glass from the overhead lights. Half of the table was considered the “shallow side” and had a fixed board covered in checkerboard material just beneath the glass. The opposite side of the table was considered the “deep side” and had a movable, checkerboard shelf that could be adjusted to be just below, 15 cm below, or 75 cm below the glass. When the shelf was adjusted to be 15 or 75 cm below the glass, square pot plastic the deep side of the table provided the illusion of depth without the risk of the bird falling. A strip of LED lights was just below and perpendicular to the glass on the deep side, in order to illuminate the glass and reduce reflection. A perch was secured in the center of the table between the division of the deep and shallow sides and 15 cm above the glass. The placement of the perch allowed the birds to sit in the center of the table and prevented the bird from receiving any tactile information about the presence of glass on the deep side; the bird could not extend its toes over the edge of the cliff or bend forward enough to peck the glass.
A platform made from poultry flooring was suspended over the deep side of the cliff at 30 cm wide and 20 cm from the perch for the small table and 37.5 cm wide and 25 cm from the perch for the large table. The bird had the option of escaping through the open ends of the canopy by either jumping from the perch to the platform or by walking across the shallow side. The depth of the cliff was adjusted prior to the start of the trial. Before each trial involving perceived depth both sides of the plexiglass was wiped down with a microfiber cloth to remove all dust and debris. The experimenter stood by the opening of the shallow side and placed the bird on perch facing the deep side of the table. The research assistant started the timer and the experimenter released their hands as soon as the bird had grasped the perch with both feet simultaneously. The experimenter remained standing at the shallow side of the table until one minute had elapsed or the bird had exited through one of the ends of the table. If after one minute the bird had not exited the apparatus, the experimenter extended their right arm towards the perch with fingers outstretched to encourage them to exit through the deep side of the table.
The trial ended after the bird had exited or after 1.5 minutes had elapsed. This procedure was repeated until the bird had completed three trials and had been presented with all three depths. Latency to jump to the platform, type of jump, and frequency of looking down, or orienting the head down, were recorded .Reliability In order to verify that behaviors were coded consistently, inter- and intra-coder agreement was calculated using Cohen’s κ. Observations were considered to be in agreement if the behaviors recorded were identical and within 2 s of each other. It was found that there was very strong agreement for intra-coder reliability with each research assistant re-coding 10 videos. For inter-coder reliability, all observers coded 5% of the videos and it was also found that they had very strong agreement . Statistical analysis Each bird was considered the experimental unit, and data were assessed for normality prior to analysis. All statistical analyses were conducted in IBM® SPSS® Statistics 27.0 with a significance level set to P < 0.05. Data are presented as the mean . A binomial logistic regression and a binomial test were run on choice to exit through the short or long arm for the unequal Y-maze configurations as well as the choice to exit left or right for the equal Y-maze configurations. The main and interaction effects of age and rearing treatment on the latency to exit the Y-maze was analyzed using a factorial ANOVA with pen as a random effect. Birds that did not exit the maze were assigned the maximum amount of time for the trial .
A generalized estimating equation was performed to investigate the main and interaction effects of age, depth and rearing treatment, as well as the random effect of pen, on crossing the visual cliff and type of movement used to cross the cliff. A factorial ANOVA was used to assess the main and interaction effects of age, depth, and rearing treatment, as well as the random effect of pen on the frequency of looking down over the cliff edge while on perch and the latency to cross the visual cliff. Birds that did not cross the cliff were assigned the maximum amount of time for the trial .In order to better understand how access to vertical structures during the pullet phase impacts the development of spatial cognition in hens, this study utilized two novel depth perception tasks: Y-maze and visual cliff test. SINGLE and MULTI birds, as well as their FLOOR reared counterparts, were shown to exit the Y-maze through the short arm significantly more frequently than what would be expected by chance. These results demonstrate that the birds were able to choose the shorter escape route when presented with the choice between 30 and 90 cm. This suggests that, regardless of rearing system, birds were able to differentiate between these two distances on a floor based task. FLOOR birds also crossed the visual cliff at the same frequency as SINGLE and MULTI reared at each cliff depth. If FLOOR birds did not adequately perceive depth, it could be expected that they would cross the cliff at the same frequency regardless of the height. However, it was found that all birds, regardless of rearing treatment, were more likely to cross at 15 cm depth than the 30 cm and 90 cm depths. Additionally, all birds had a significantly shorter latency when crossing at the 15 cm depth when compared to the greater depths. These results demonstrate that the birds were responding differently to variations in depth, suggesting that FLOOR birds do not have a deficit in the ability to perceive depth when compared to SINGLE and MULTI birds. This is an intriguing addition to the current evidence that access to vertical space at a young age improves spatial memory and navigation in hens . Although these aspects of spatial cognition are affected by access to perches and platforms during development, square pots for plants depth perception appears to be unaffected.This study provides evidence that the development of accurate depth perception does not require specific visual experience, such as exposure to depth and height, during rearing. This is supported by studies with one-day old chicks and four-day old chicks without exposure to vertical structures which found these animals already possessed the ability to perceive depth. Gibson and Walk preformed a cross species comparison of depth perception abilities by testing infant rat pups, kittens, chicks, lambs, goat kids, and piglets on a visual cliff. They found that all animals tested, including animals that were less than one day of age, demonstrated avoidance of the deep side of the visual cliff, akin to that of their adult counterparts. Gibson and Walk concluded that depth perception abilities are most likely innate, however did not exclude the chance that learning may play a role to some degree. The theory that depth perception is an innate skill was challenged by Tallarico and Ferrell when they raised chicks on either the deep side or the shallow side of a visual cliff for the first 30-40 hours of life. When placed in the center of a visual cliff with both deep side and shallow sides present, chicks reared in the deep side environment were significantly more likely to cross to the deep side of the cliff than those reared on the shallow side.
Tallarico and Ferrell used this as evidence that there is not innate avoidance of a drop off or cliff. However, chicks reared in the deep side environment may have learned that the cliff was not a real threat and that there was a barrier preventing them from falling. Instead, evidence from previous research as well as the current study suggests that depth perception is an innate ability and rearing chicks on the floor does not prevent the proper development of depth perception. Despite the same ability to perceive depth between the three treatments, differences in frequency of crossing the visual cliff was observed between the FLOOR, SINGLE, and MULTIbirds. The FLOOR birds were significantly less likely to cross the visual cliff than the other two treatments at 8 and 16 weeks of age. This difference dissipated at week 30, when all birds, regardless of rearing treatment, crossed the visual cliff at the same rate. This suggests that something inhibited the FLOOR birds from crossing the visual cliff at a younger age when housed in the treatment pen and that this reversed after the FLOOR birds were exposed to their vertically complex adult housing after 16 weeks. One possibility is that MULTI and SINGLE birds may have been more comfortable with the platform as a landing surface due to their experience with poultry flooring in their home pens. However, the appearance of the poultry flooring in the visual cliff and rearing pen systems was different in color and size. The color of the poultry flooring in the MULTI and SINGLE rearing treatment pens was black as opposed to the white poultry flooring used for the visual cliff. Also, the platform was a much smaller landing surface than the large platforms used in the MULTI and SINGLE rearing systems as well as the adult multi-tier aviary system. Another possibility is that FLOOR birds were less likely to cross due to lack of experience with tall vertical structures such as perches and platforms above 30 cm in height. This lack of experience with vertical structures may prevent hens reared on the floor from readily crossing between perches and platforms when first introduced to a more vertically complex environment. FLOOR birds at 8 weeks of age also showed significantly longer latencies to cross the visual cliff than SINGLE and MULTI birds at 8 weeks of age. This further supports our suggestion that FLOOR reared birds were tentative when using vertical structures prior to being transitioned into their adult housing. However, it is interesting that FLOOR birds at 16 weeks did not have significantly longer latency than the other rearing treatments. Meaning that FLOOR pullets at 16 weeks of age were less hesitant than their 8 week old counterparts to utilize vertical structures. Across all treatments, birds at 16 weeks of age were less likely to cross the visual cliff than at 8 or 30 weeks old. Additionally, all birds exhibited longer latencies to exit the Y-maze and cross the visual cliff at this intermediate time point when compared to the other time points. The visual cliff was smaller at the 8 week time point and was increased in size by 25% for the 16 and 30 week time points. The size of the visual cliff was increased after the first time point due to the increased size of the birds. It is important to note that the frequency of crossing and latencies to cross were not significantly different between the 8 week and 30 week old birds despite the table being the same size at 30 weeks as it was at 16 weeks. Also, the birds did not cross the cliff at 16 weeks of age significantly differently than at the other time points, demonstrating that they did not need to jump or fly to reach the platform. Therefore, it is unlikely that the size difference of the visual cliff contributed to the avoidance of crossing and longer latencies observed at the 16 week time point. It would be expected that if the increase in size of the visual cliff increased the physical difficulty of the task, the birds would have no longer been able to consistently step from the perch to the platform for the visual cliff.